Voices of Search Podcast: Summary

Episode Title: G2 Data: 1/2 of Global Software Buyers Now Start Search on AI Chatbots Instead of Google
Date: February 9, 2026
Host: Jordan Cooney
Guest: Tim Sanders, Chief Innovation Officer at G2

Overview

This episode tackles the rapid transformation in B2B software buying behaviors as AI search and large language models (LLMs) upend traditional, Google-centric approaches. Drawing from G2’s massive data set and Tim Sanders’ deep industry perspective, the discussion explores how buyers now prefer AI chatbots to conduct research, what this means for SEO and content strategies, and how companies must adapt to the LLM-first era to maintain (and measure) discoverability, relevance, and buyer trust.

Key Discussion Points and Insights

1. The Death of “Google First” for Software Buyers

(00:44 – 02:01)

• 79% of global B2B buyers report that AI search has fundamentally changed how they conduct research.
• Key Data: 8 in 10 buyers gather, compare, and shortlist vendors using AI before clicking traditional results.
• Buyers are now starting with AI chatbots, not Google searches—putting classic SEO strategies at risk.

Quote:
“Today’s buyers are not starting with search. They're focused on AI chat. That’s the key result in success.”
—Jordan Cooney (00:53)

2. Tim Sanders’ History in Search & Patterns of Disruption

(02:02 – 05:42)

• Tim recounts early days at Yahoo and watching Google’s PageRank upend everything from print to digital.
• Key lesson: Search strategy needs to focus on durable, long-term practices—not chase temporary tricks.
• With AI-centric search, reliable analytics/measurement frameworks haven’t fully caught up yet.

Quote:
“I tended to spend most of my time on... the things that don’t change over time. Like, what are the long-standing practices you can have from a site structure and content standpoint that could create continuity in your search?”
—Tim Sanders (04:32)

3. New Buyer Journey: From Reference to Inference

(06:18 – 08:31)

• Classic search: Buyer goes to Google (the “great librarian”), scans blue links, manually builds a shortlist (5–12 hours).
• AI search: Share your goal/problem, and get an inferred shortlist from an LLM in minutes, not hours.
• The bottleneck is no longer gathering options but verifying them.

Quote:
“The software buyer... has moved from reference to inference.”
—Tim Sanders (07:09)

4. Hallucinations: Human vs. Machine Errors

(08:31 – 12:40)

• While early LLMs hallucinated more, error rates have plummeted (now often 3–5% or less).
• Human “confident errors” (i.e., believing untrue things) are often higher and rising.
• Effective use of prompting and adversarial querying can all but eliminate LLM hallucinations in practical workflows.
• Hallucinations, both human and machine, are sometimes a feature, sparking creativity and breakthroughs.

Quote:
“There’s an inflection point coming... where the hallucination rate of a human is going to be statistically higher than a machine—and not by a small margin.”
—Tim Sanders (10:38)

5. G2’s Role in the Verification Layer

(13:31 – 20:08)

• As LLMs generate shortlists, buyers still click through to citation sources for verification.
• G2’s “Best of” guides and verified reviews are heavily cited because they fit LLM prompts (“best three CRMs for…”).
• G2’s rigorous verification (screenshots, LinkedIn, etc.) creates trust—attracting more AI referral traffic.

Data:
“In our August survey—same question: 50% [of B2B buyers start on a chatbot instead of Google].”
—Tim Sanders (13:31)

6. Verification Clicks & Quality of Traffic

(20:08 – 24:44)

• Not all citations bring traffic. Commercial citations (e.g., recommendations for specific products) drive much higher clickthroughs (3–15%) than background ones.
• LLM-driven site visitors convert at higher rates because they arrive with greater intent and further along cognitively in the buying cycle.

Quote:
“The traffic that comes to us from language models is much higher conviction. They’re much further down the cognitive process.”
—Tim Sanders (22:03)

7. Rethinking Content, Measurement & Optimization for AEO/Geo

(26:12 – 33:53)

• New focus on user’s “cognitive age” vs. “calendar age” in the funnel—AI users arrive more decided, less fatigued.
• Calculating expected value (EV) is vital—should we ungate content for LLM visibility or gate for lead gen? Use EV to decide.
• AI models struggle with content in PDFs/gated assets; republishing as HTML/Markdown boosts LLM inclusion.

Quote:
“You have to use the concept of expected value now when you make fundamental marketing decisions.”
—Tim Sanders (31:53)

8. Shifts in Buying Power and Benchmarks

(34:51 – 39:16)

• IT is rapidly consolidating power as the primary AI “buyer.” Review recency, richness, and technical benchmarks (hallucination rate, reliability) now matter more than classic NPS.
• Review token length/story arc also improves LLM recognition and value.

9. Answer-Shaped, Natural Language Content Wins

(40:23 – 42:54)

• Video and podcasts (with transcripts) excel because they’re naturally “answer-shaped” and less corporate, preferred by LLMs for verification.
• Ensuring transcripts are visible to bots is essential.

Quote:
“YouTube and podcast are a forcing function for you... to refactor content to be answer shaped instead of message shaped.”
—Tim Sanders (42:37)

10. The Changing Discovery Landscape: Multipolar AI Platforms

(42:59 – 44:15)

• It’s not just Google/Gemini—buyers also use ChatGPT, Claude, and Copilot. For B2B, Copilot shouldn’t be ignored.
• Marketers must treat each as a unique "analyst" and understand their citation and recommendation logic.

11. Genuine Marketing in the Regrettable-Purchase Era

(45:49 – 48:08)

• Any attempt to “game” reviews or community content will be quickly detected; AI closes loopholes rapidly.
• Best marketing aligns with genuine empathy and buyer trust; focus on helping the right buyer avoid regret.
• SEO fundamentals increasingly apply to AEO—a new swim lane, not a replacement.

Quote:
“If you market from a place of empathy—that what you’re trying to do is help the right buyer find the right product and never regret the purchase—you’re aligning with something that will probably be as endearing... as PageRank.”
—Tim Sanders (46:09)

Notable Quotes & Memorable Moments

• “The hallucination rate of a human is going to be statistically higher than a machine.” —Tim Sanders (10:38)
• "You have to use the concept of expected value [EV]... when you make fundamental marketing decisions." —Tim Sanders (31:53)
• “If you rely too much on keywords in AEO, you’re going to miss the mark—and worst of all, create false positives.” —Tim Sanders (50:18)
• "Think of Copilot like an analyst... respect it, study it, understand how it lives and breathes." —Tim Sanders (43:28)
• “Verifier’s Law: Anything that's easy to verify will be conquered by AI. Love it. So regrettable purchases, easy to verify. Just tell the truth.” —Tim Sanders (47:13)

Key Timestamps for Major Segments

• 00:44 – G2 research: AI chatbots overtaking Google as starting point.
• 02:25 – Tim’s journey from Yahoo to AI search, lessons learned.
• 06:18 – The buyer’s journey: reference vs. inference.
• 08:51 – On hallucinations: machine vs. human, and reducing errors.
• 13:31 – Citation value, LLMs, and G2’s verification layer.
• 19:20 – Click-through rates from AI citations, trust and verification.
• 26:12 – Buyer’s “cognitive age” vs. calendar age; impact on marketing.
• 31:53 – Using expected value (EV) in content/marketing decisions.
• 34:51 – IT’s central role in AI purchases, changing review value.
• 40:23 – Why “answer-shaped” (natural language) content wins.
• 42:59 – Multipolar AI search platforms: Copilot, Claude, ChatGPT, Gemini.
• 45:49 – Gaming reviews and “regrettable purchases.”
• 48:19 to End – Lightning round: surprising data, prompts vs. keywords, markdown key takeaways, future content best practices for AI.

Actionable Takeaways

• Refocus SEO: Double down on answer-shaped, factual, and recent reviews and content—especially in HTML/Markdown—to be indexed and cited by LLMs.
• Audit for AI crawlability: Remove unnecessary gating and PDFs, and use expected value calculations before changing legacy practices.
• Embrace Multimedia: Video and podcasts with accessible transcripts are rising signals for LLM discovery.
• Stay Authentic: Focus on genuine, verified customer experience—models will rapidly close loopholes for manipulation.
• Measure What Matters: Track not just traffic, but calibration of outcomes, post-buy events, and evolving buyer personas (IT now dominates for AI/software).
• Don’t Ignore New AI Channels: Tailor approaches for Copilot, Gemini, Claude, and others. Each AI channel is an “analyst” with unique traits.

Final Thoughts

This episode reaffirms that the AI search era is not a distant horizon—it’s here, and it’s already transforming B2B software discovery, evaluation, and conversion. Legacy SEO is still relevant but needs to evolve rapidly. Marketers must master both new technical tactics and the strategic discipline of storytelling, transparency, and precise targeting to prosper in an LLM-first world.

For deeper insights, connect with Tim Sanders via LinkedIn or visit voicesofsearch.com.